Skip to contentSkip to conversation

fpga: Remove lattice and ACEX1 support

PATCH Superseded Michal Simek 4 participants 13 messages Jul 25, 2025 – Jul 26, 2025

Subsystems (2)

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore

Michal Simek

author maintainer
#0 lore raw

cover fpga: Remove lattice and ACEX1 support

Hi,

when fpga drivers were enabled via sandbox some coverity issues came and make no sense to fix them when there is no user of these drivers that's why it is easier just remove them completely.

Another candidate for removal is FPGA_CYCLON2 which is enabled only by mx53cx9020_defconfig.
Patrick: Is this board still maintained?

Thanks,
Michal

Michal Simek (3):
fpga: xilinx: Check valid desc structure
fpga: Remove ancient ACEX1K support
fpga: lattice: Remove unused support

 configs/sandbox_defconfig |    2 -
 drivers/fpga/ACEX1K.c     |  245 ---
 drivers/fpga/Kconfig      |   12 -
 drivers/fpga/Makefile     |    2 -
 drivers/fpga/altera.c     |    5 +-
 drivers/fpga/fpga.c       |   23 -
 drivers/fpga/ivm_core.c   | 3149 -------------------------------------
 drivers/fpga/lattice.c    |  379 -----
 drivers/fpga/xilinx.c     |    3 +
 include/ACEX1K.h          |   49 -
 include/fpga.h            |    1 -
 include/lattice.h         |  298 ----
 12 files changed, 4 insertions(+), 4164 deletions(-)
 delete mode 100644 drivers/fpga/ACEX1K.c
 delete mode 100644 drivers/fpga/ivm_core.c
 delete mode 100644 drivers/fpga/lattice.c
 delete mode 100644 include/lattice.h
Signature · 4 lines
2.43.0

base-commit: 021783860f7e628f7c4e0c101707cd4250e6d61f
branch: debian-sent3
author maintainer
#1 lore raw

PATCH 1/3 fpga: xilinx: Check valid desc structure

FPGA validation can fail and return value needs to be checked.

Addresses-Coverity-ID: CID 583150: Null pointer dereferences (NULL_RETURNS)

---

 drivers/fpga/xilinx.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/fpga/xilinx.c +3 -0
index 28c68faba55f..25b348648eff 100644
@@ -50,6 +50,9 @@ int fpga_loadbitstream(int devnum, char *fpgadata, size_t size,
 	dataptr = (unsigned char *)fpgadata;
 	/* Find out fpga_description */
 	desc = fpga_validate(devnum, dataptr, 0);
+	if (!desc)
+		return FPGA_FAIL;
+
 	/* Assign xilinx device description */
 	xdesc = desc->devdesc;
 
Signature · 1 line
2.43.0
author maintainer
#2 lore raw

PATCH 2/3 fpga: Remove ancient ACEX1K support

In reply to Michal Simek

Coverity (CID 583149) reports issue on code which is not enabled by any real platform that's why remove it completely.

---

 configs/sandbox_defconfig |   1 -
 drivers/fpga/ACEX1K.c     | 245 --------------------------------------
 drivers/fpga/Kconfig      |   6 -
 drivers/fpga/Makefile     |   1 -
 drivers/fpga/altera.c     |   5 +-
 include/ACEX1K.h          |  49 --------
 6 files changed, 1 insertion(+), 306 deletions(-)
 delete mode 100644 drivers/fpga/ACEX1K.c
configs/sandbox_defconfig +0 -1
index 0c030f4a7925..96546cc0bcfe 100644
@@ -206,7 +206,6 @@ CONFIG_ARM_FFA_TRANSPORT=y
 CONFIG_FPGA_ALTERA=y
 CONFIG_FPGA_STRATIX_II=y
 CONFIG_FPGA_STRATIX_V=y
-CONFIG_FPGA_ACEX1K=y
 CONFIG_FPGA_CYCLON2=y
 CONFIG_FPGA_LATTICE=y
 CONFIG_FPGA_XILINX=y
drivers/fpga/ACEX1K.c deleted +0 -245

245 lines not on this page · raw message

drivers/fpga/Kconfig +0 -6
index 9456ca3149ae..18e7d7c99a9a 100644
@@ -34,12 +34,6 @@ config FPGA_STRATIX_V
 	help
 	  Say Y here to enable the Altera Stratix V FPGA specific driver.
 
-config FPGA_ACEX1K
-	bool "Enable Altera ACEX 1K driver"
-	depends on FPGA_ALTERA
-	help
-	  Say Y here to enable the Altera ACEX 1K FPGA specific driver.
-
 config FPGA_CYCLON2
 	bool "Enable Altera FPGA driver for Cyclone II"
 	depends on FPGA_ALTERA
drivers/fpga/Makefile +0 -1
index 610c168fc35c..6c779d5b5df6 100644
@@ -17,7 +17,6 @@ obj-$(CONFIG_FPGA_XILINX) += xilinx.o
 obj-$(CONFIG_FPGA_LATTICE) += ivm_core.o lattice.o
 ifdef CONFIG_FPGA_ALTERA
 obj-y += altera.o
-obj-$(CONFIG_FPGA_ACEX1K) += ACEX1K.o
 obj-$(CONFIG_FPGA_CYCLON2) += cyclon2.o
 obj-$(CONFIG_FPGA_INTEL_SDM_MAILBOX) += intel_sdm_mb.o
 obj-$(CONFIG_FPGA_STRATIX_II) += stratixII.o
drivers/fpga/altera.c +1 -4
index 64fda3a307c7..4a9aa74357ec 100644
@@ -28,10 +28,7 @@ static const struct altera_fpga {
 	int			(*dump)(Altera_desc *, const void *, size_t);
 	int			(*info)(Altera_desc *);
 } altera_fpga[] = {
-#if defined(CONFIG_FPGA_ACEX1K)
-	{ Altera_ACEX1K, "ACEX1K", ACEX1K_load, ACEX1K_dump, ACEX1K_info },
-	{ Altera_CYC2,   "ACEX1K", ACEX1K_load, ACEX1K_dump, ACEX1K_info },
-#elif defined(CONFIG_FPGA_CYCLON2)
+#if defined(CONFIG_FPGA_CYCLON2)
 	{ Altera_ACEX1K, "CycloneII", CYC2_load, CYC2_dump, CYC2_info },
 	{ Altera_CYC2,   "CycloneII", CYC2_load, CYC2_dump, CYC2_info },
 #endif
include/ACEX1K.h +0 -49

63 lines not on this page · raw message

Signature · 1 line
2.43.0
#3 lore raw

Re: [PATCH 2/3] fpga: Remove ancient ACEX1K support

Hello Michal,

Am Fri, Jul 25, 2025 at 01:25:53PM +0200 schrieb Michal Simek:

Quoted 328 lines

Coverity (CID 583149) reports issue on code which is not enabled by any real platform that's why remove it completely.

---

 configs/sandbox_defconfig |   1 -
 drivers/fpga/ACEX1K.c     | 245 --------------------------------------
 drivers/fpga/Kconfig      |   6 -
 drivers/fpga/Makefile     |   1 -
 drivers/fpga/altera.c     |   5 +-
 include/ACEX1K.h          |  49 --------
 6 files changed, 1 insertion(+), 306 deletions(-)
 delete mode 100644 drivers/fpga/ACEX1K.c
configs/sandbox_defconfig +0 -1
index 0c030f4a7925..96546cc0bcfe 100644
@@ -206,7 +206,6 @@ CONFIG_ARM_FFA_TRANSPORT=y
 CONFIG_FPGA_ALTERA=y
 CONFIG_FPGA_STRATIX_II=y
 CONFIG_FPGA_STRATIX_V=y
-CONFIG_FPGA_ACEX1K=y
 CONFIG_FPGA_CYCLON2=y
 CONFIG_FPGA_LATTICE=y
 CONFIG_FPGA_XILINX=y
drivers/fpga/ACEX1K.c deleted +0 -245
deleted file mode 100644
index e1514fc56d04..000000000000
@@ -1,245 +0,0 @@
-// SPDX-License-Identifier: GPL-2.0+
-/*
- * (C) Copyright 2003
- * Steven Scholz, imc Measurement & Control, [email protected]
- *
- * (C) Copyright 2002
- * Rich Ireland, Enterasys Networks, [email protected].
- */
-
-#define LOG_CATEGORY UCLASS_FPGA
-
-#include <config.h>		/* core U-Boot definitions */
-#include <console.h>
-#include <log.h>
-#include <ACEX1K.h>		/* ACEX device family */
-#include <linux/delay.h>
-#include <time.h>
-
-/* Note: The assumption is that we cannot possibly run fast enough to
- * overrun the device (the Slave Parallel mode can free run at 50MHz).
- * If there is a need to operate slower, define CFG_FPGA_DELAY in
- * the board config file to slow things down.
- */
-#ifndef CFG_FPGA_DELAY
-#define CFG_FPGA_DELAY()
-#endif
-
-#ifndef CFG_SYS_FPGA_WAIT
-#define CFG_SYS_FPGA_WAIT CONFIG_SYS_HZ/10		/* 100 ms */
-#endif
-
-static int ACEX1K_ps_load(Altera_desc *desc, const void *buf, size_t bsize);
-static int ACEX1K_ps_dump(Altera_desc *desc, const void *buf, size_t bsize);
-/* static int ACEX1K_ps_info(Altera_desc *desc); */
-
-/* ------------------------------------------------------------------------- */
-/* ACEX1K Generic Implementation */
-int ACEX1K_load(Altera_desc *desc, const void *buf, size_t bsize)
-{
-	int ret_val = FPGA_FAIL;
-
-	switch (desc->iface) {
-	case passive_serial:
-		log_debug("Launching Passive Serial Loader\n");
-		ret_val = ACEX1K_ps_load (desc, buf, bsize);
-		break;
-
-		/* Add new interface types here */
-
-	default:
-		printf ("%s: Unsupported interface type, %d\n",
-				__FUNCTION__, desc->iface);
-	}
-
-	return ret_val;
-}
-
-int ACEX1K_dump(Altera_desc *desc, const void *buf, size_t bsize)
-{
-	int ret_val = FPGA_FAIL;
-
-	switch (desc->iface) {
-	case passive_serial:
-		log_debug("Launching Passive Serial Dump\n");
-		ret_val = ACEX1K_ps_dump (desc, buf, bsize);
-		break;
-
-		/* Add new interface types here */
-
-	default:
-		printf ("%s: Unsupported interface type, %d\n",
-				__FUNCTION__, desc->iface);
-	}
-
-	return ret_val;
-}
-
-int ACEX1K_info( Altera_desc *desc )
-{
-	return FPGA_SUCCESS;
-}
-
-/* ------------------------------------------------------------------------- */
-/* ACEX1K Passive Serial Generic Implementation                                  */
-
-static int ACEX1K_ps_load(Altera_desc *desc, const void *buf, size_t bsize)
-{
-	int ret_val = FPGA_FAIL;	/* assume the worst */
-	Altera_ACEX1K_Passive_Serial_fns *fn = desc->iface_fns;
-	int i;
-
-	log_debug("start with interface functions @ 0x%p\n", fn);
-
-	if (fn) {
-		size_t bytecount = 0;
-		unsigned char *data = (unsigned char *) buf;
-		int cookie = desc->cookie;	/* make a local copy */
-		unsigned long ts;		/* timestamp */
-
-		log_debug("Function Table:\n"
-			  "ptr:\t0x%p\n"
-			  "struct: 0x%p\n"
-			  "config:\t0x%p\n"
-			  "status:\t0x%p\n"
-			  "clk:\t0x%p\n"
-			  "data:\t0x%p\n"
-			  "done:\t0x%p\n\n",
-			  &fn, fn, fn->config, fn->status,
-			  fn->clk, fn->data, fn->done);
-#ifdef CONFIG_SYS_FPGA_PROG_FEEDBACK
-		printf ("Loading FPGA Device %d...", cookie);
-#endif
-
-		/*
-		 * Run the pre configuration function if there is one.
-		 */
-		if (*fn->pre) {
-			(*fn->pre) (cookie);
-		}
-
-		/* Establish the initial state */
-		(*fn->config) (true, true, cookie);	/* Assert nCONFIG */
-
-		udelay(2);		/* T_cfg > 2us	*/
-
-		/* nSTATUS should be asserted now */
-		(*fn->done) (cookie);
-		if ( !(*fn->status) (cookie) ) {
-			puts ("** nSTATUS is not asserted.\n");
-			(*fn->abort) (cookie);
-			return FPGA_FAIL;
-		}
-
-		(*fn->config) (false, true, cookie);	/* Deassert nCONFIG */
-		udelay(2);		/* T_cf2st1 < 4us	*/
-
-		/* Wait for nSTATUS to be released (i.e. deasserted) */
-		ts = get_timer (0);		/* get current time */
-		do {
-			CFG_FPGA_DELAY ();
-			if (get_timer (ts) > CFG_SYS_FPGA_WAIT) {	/* check the time */
-				puts ("** Timeout waiting for STATUS to go high.\n");
-				(*fn->abort) (cookie);
-				return FPGA_FAIL;
-			}
-			(*fn->done) (cookie);
-		} while ((*fn->status) (cookie));
-
-		/* Get ready for the burn */
-		CFG_FPGA_DELAY ();
-
-		/* Load the data */
-		while (bytecount < bsize) {
-			unsigned char val=0;
-#ifdef CONFIG_SYS_FPGA_CHECK_CTRLC
-			if (ctrlc ()) {
-				(*fn->abort) (cookie);
-				return FPGA_FAIL;
-			}
-#endif
-			/* Altera detects an error if INIT goes low (active)
-			   while DONE is low (inactive) */
-#if 0 /* not yet implemented */
-			if ((*fn->done) (cookie) == 0 && (*fn->init) (cookie)) {
-				puts ("** CRC error during FPGA load.\n");
-				(*fn->abort) (cookie);
-				return (FPGA_FAIL);
-			}
-#endif
-			val = data [bytecount ++ ];
-			i = 8;
-			do {
-				/* Deassert the clock */
-				(*fn->clk) (false, true, cookie);
-				CFG_FPGA_DELAY ();
-				/* Write data */
-				(*fn->data) ((val & 0x01), true, cookie);
-				CFG_FPGA_DELAY ();
-				/* Assert the clock */
-				(*fn->clk) (true, true, cookie);
-				CFG_FPGA_DELAY ();
-				val >>= 1;
-				i --;
-			} while (i > 0);
-
-#ifdef CONFIG_SYS_FPGA_PROG_FEEDBACK
-			if (bytecount % (bsize / 40) == 0)
-				putc ('.');		/* let them know we are alive */
-#endif
-		}
-
-		CFG_FPGA_DELAY ();
-
-#ifdef CONFIG_SYS_FPGA_PROG_FEEDBACK
-		putc (' ');			/* terminate the dotted line */
-#endif
-
-	/*
-	 * Checking FPGA's CONF_DONE signal - correctly booted ?
-	 */
-
-	if ( ! (*fn->done) (cookie) ) {
-		puts ("** Booting failed! CONF_DONE is still deasserted.\n");
-		(*fn->abort) (cookie);
-		return (FPGA_FAIL);
-	}
-
-	/*
-	 * "DCLK must be clocked an additional 10 times fpr ACEX 1K..."
-	 */
-
-	for (i = 0; i < 12; i++) {
-		CFG_FPGA_DELAY ();
-		(*fn->clk) (true, true, cookie);	/* Assert the clock pin */
-		CFG_FPGA_DELAY ();
-		(*fn->clk) (false, true, cookie);	/* Deassert the clock pin */
-	}
-
-	ret_val = FPGA_SUCCESS;
-
-#ifdef CONFIG_SYS_FPGA_PROG_FEEDBACK
-		if (ret_val == FPGA_SUCCESS) {
-			puts ("Done.\n");
-		}
-		else {
-			puts ("Fail.\n");
-		}
-#endif
-	(*fn->post) (cookie);
-
-	} else {
-		printf ("%s: NULL Interface function table!\n", __FUNCTION__);
-	}
-
-	return ret_val;
-}
-
-static int ACEX1K_ps_dump(Altera_desc *desc, const void *buf, size_t bsize)
-{
-	/* Readback is only available through the Slave Parallel and         */
-	/* boundary-scan interfaces.                                         */
-	printf ("%s: Passive Serial Dumping is unavailable\n",
-			__FUNCTION__);
-	return FPGA_FAIL;
-}
drivers/fpga/Kconfig +0 -6
index 9456ca3149ae..18e7d7c99a9a 100644
@@ -34,12 +34,6 @@ config FPGA_STRATIX_V
 	help
 	  Say Y here to enable the Altera Stratix V FPGA specific driver.
 
-config FPGA_ACEX1K
-	bool "Enable Altera ACEX 1K driver"
-	depends on FPGA_ALTERA
-	help
-	  Say Y here to enable the Altera ACEX 1K FPGA specific driver.
-
 config FPGA_CYCLON2
 	bool "Enable Altera FPGA driver for Cyclone II"
 	depends on FPGA_ALTERA
drivers/fpga/Makefile +0 -1
index 610c168fc35c..6c779d5b5df6 100644
@@ -17,7 +17,6 @@ obj-$(CONFIG_FPGA_XILINX) += xilinx.o
 obj-$(CONFIG_FPGA_LATTICE) += ivm_core.o lattice.o
 ifdef CONFIG_FPGA_ALTERA
 obj-y += altera.o
-obj-$(CONFIG_FPGA_ACEX1K) += ACEX1K.o
 obj-$(CONFIG_FPGA_CYCLON2) += cyclon2.o
 obj-$(CONFIG_FPGA_INTEL_SDM_MAILBOX) += intel_sdm_mb.o
 obj-$(CONFIG_FPGA_STRATIX_II) += stratixII.o
drivers/fpga/altera.c +1 -4
index 64fda3a307c7..4a9aa74357ec 100644
@@ -28,10 +28,7 @@ static const struct altera_fpga {
 	int			(*dump)(Altera_desc *, const void *, size_t);
 	int			(*info)(Altera_desc *);
 } altera_fpga[] = {
-#if defined(CONFIG_FPGA_ACEX1K)
-	{ Altera_ACEX1K, "ACEX1K", ACEX1K_load, ACEX1K_dump, ACEX1K_info },
-	{ Altera_CYC2,   "ACEX1K", ACEX1K_load, ACEX1K_dump, ACEX1K_info },
-#elif defined(CONFIG_FPGA_CYCLON2)
+#if defined(CONFIG_FPGA_CYCLON2)
 	{ Altera_ACEX1K, "CycloneII", CYC2_load, CYC2_dump, CYC2_info },
 	{ Altera_CYC2,   "CycloneII", CYC2_load, CYC2_dump, CYC2_info },
 #endif
include/ACEX1K.h +0 -0
index 7c5253c66cca..44d1b9765657 100644

include/ACEX1K.h line 17

Quoted 10 lines
include/ACEX1K.h +0 -4
@@ -12,26 +12,10 @@
 
 #include <altera.h>
 
-extern int ACEX1K_load(Altera_desc *desc, const void *image, size_t size);
-extern int ACEX1K_dump(Altera_desc *desc, const void *buf, size_t bsize);
-extern int ACEX1K_info(Altera_desc *desc);
-
 extern int CYC2_load(Altera_desc *desc, const void *image, size_t size);
 extern int CYC2_dump(Altera_desc *desc, const void *buf, size_t bsize);
 extern int CYC2_info(Altera_desc *desc);

Maybe CYC2_load() etc. can be moved to a different header file like altera.h for example?

Quoted 36 lines
include/ACEX1K.h +0 -28
-/* Slave Serial Implementation function table */
-typedef struct {
-	Altera_pre_fn		pre;
-	Altera_config_fn	config;
-	Altera_clk_fn		clk;
-	Altera_status_fn	status;
-	Altera_done_fn		done;
-	Altera_data_fn		data;
-	Altera_abort_fn		abort;
-	Altera_post_fn		post;
-} Altera_ACEX1K_Passive_Serial_fns;
-
 /* Slave Serial Implementation function table */
 typedef struct {
 	Altera_pre_fn		pre;
@@ -43,37 +27,4 @@ typedef struct {
 	Altera_post_fn		post;
 } Altera_CYC2_Passive_Serial_fns;
 
-/* Device Image Sizes
- *********************************************************************/
-/* ACEX1K */
-/* FIXME: Which size do we mean?
- * Datasheet says 1337000/8=167125Bytes,
- * Filesize of an *.rbf file is 166965 Bytes
- */
-#if 0
-#define Altera_EP1K100_SIZE	1337000/8	/* 167125 Bytes */
-#endif
-#define Altera_EP1K100_SIZE	(166965*8)
-
-#define Altera_EP2C8_SIZE	247942
-#define Altera_EP2C20_SIZE	586562
-#define Altera_EP2C35_SIZE	883905
-#define Altera_EP3C5_SIZE	368011		/* .rbf size in bytes */

include/ACEX1K.h typedef struct { · line 71

Quoted 10 lines
include/ACEX1K.h +0 -10
-
-#define ALTERA_EP4CE6_SIZE	368011		/* 2944088 Bits */
-#define ALTERA_EP4CE10_SIZE	368011		/* 2944088 Bits */
-#define ALTERA_EP4CE15_SIZE	510856		/* 4086848 Bits */
-#define ALTERA_EP4CE22_SIZE	718569		/* 5748552 Bits */
-#define ALTERA_EP4CE30_SIZE	1191788		/* 9534304 Bits */
-#define ALTERA_EP4CE40_SIZE	1191788		/* 9534304 Bits */
-#define ALTERA_EP4CE55_SIZE	1861195		/* 14889560 Bits */
-#define ALTERA_EP4CE75_SIZE	2495719		/* 19965752 Bits */
-#define ALTERA_EP4CE115_SIZE	3571462		/* 28571696 Bits */

Can we move these definitions for device image sizes to altera.h instead removing them? I added some of those with commit 3b2a595fc659 ("fpga: altera: Add some more device sizes") back in 2019, and I'm using some of them in downstream code/dts.

Greets
Alex

Quoted 11 lines
include/ACEX1K.h +0 -7
-
-/* Descriptor Macros
- *********************************************************************/
-/* ACEX1K devices */
-#define Altera_EP1K100_DESC(iface, fn_table, cookie) \
-{ Altera_ACEX1K, iface, Altera_EP1K100_SIZE, fn_table, cookie }
-
 #endif /* _ACEX1K_H_ */
Signature · 1 line
2.43.0
author maintainer
#4 lore raw

Re: [PATCH 2/3] fpga: Remove ancient ACEX1K support

Quoted 388 lines · Alexander Dahl

Hello Michal,

Am Fri, Jul 25, 2025 at 01:25:53PM +0200 schrieb Michal Simek:

Quoted 328 lines

Coverity (CID 583149) reports issue on code which is not enabled by any real platform that's why remove it completely.

---

  configs/sandbox_defconfig |   1 -
  drivers/fpga/ACEX1K.c     | 245 --------------------------------------
  drivers/fpga/Kconfig      |   6 -
  drivers/fpga/Makefile     |   1 -
  drivers/fpga/altera.c     |   5 +-
  include/ACEX1K.h          |  49 --------
  6 files changed, 1 insertion(+), 306 deletions(-)

delete mode 100644 drivers/fpga/ACEX1K.c

configs/sandbox_defconfig +0 -1
index 0c030f4a7925..96546cc0bcfe 100644
@@ -206,7 +206,6 @@ CONFIG_ARM_FFA_TRANSPORT=y
  CONFIG_FPGA_ALTERA=y
  CONFIG_FPGA_STRATIX_II=y
  CONFIG_FPGA_STRATIX_V=y
-CONFIG_FPGA_ACEX1K=y
  CONFIG_FPGA_CYCLON2=y
  CONFIG_FPGA_LATTICE=y
  CONFIG_FPGA_XILINX=y
drivers/fpga/ACEX1K.c deleted +0 -245
deleted file mode 100644
index e1514fc56d04..000000000000
@@ -1,245 +0,0 @@
-// SPDX-License-Identifier: GPL-2.0+
-/*
- * (C) Copyright 2003
- * Steven Scholz, imc Measurement & Control, [email protected]
- *
- * (C) Copyright 2002
- * Rich Ireland, Enterasys Networks, [email protected].
- */
-
-#define LOG_CATEGORY UCLASS_FPGA
-
-#include <config.h>		/* core U-Boot definitions */
-#include <console.h>
-#include <log.h>
-#include <ACEX1K.h>		/* ACEX device family */
-#include <linux/delay.h>
-#include <time.h>
-
-/* Note: The assumption is that we cannot possibly run fast enough to
- * overrun the device (the Slave Parallel mode can free run at 50MHz).
- * If there is a need to operate slower, define CFG_FPGA_DELAY in
- * the board config file to slow things down.
- */
-#ifndef CFG_FPGA_DELAY
-#define CFG_FPGA_DELAY()
-#endif
-
-#ifndef CFG_SYS_FPGA_WAIT
-#define CFG_SYS_FPGA_WAIT CONFIG_SYS_HZ/10		/* 100 ms */
-#endif
-
-static int ACEX1K_ps_load(Altera_desc *desc, const void *buf, size_t bsize);
-static int ACEX1K_ps_dump(Altera_desc *desc, const void *buf, size_t bsize);
-/* static int ACEX1K_ps_info(Altera_desc *desc); */
-
-/* ------------------------------------------------------------------------- */
-/* ACEX1K Generic Implementation */
-int ACEX1K_load(Altera_desc *desc, const void *buf, size_t bsize)
-{
-	int ret_val = FPGA_FAIL;
-
-	switch (desc->iface) {
-	case passive_serial:
-		log_debug("Launching Passive Serial Loader\n");
-		ret_val = ACEX1K_ps_load (desc, buf, bsize);
-		break;
-
-		/* Add new interface types here */
-
-	default:
-		printf ("%s: Unsupported interface type, %d\n",
-				__FUNCTION__, desc->iface);
-	}
-
-	return ret_val;
-}
-
-int ACEX1K_dump(Altera_desc *desc, const void *buf, size_t bsize)
-{
-	int ret_val = FPGA_FAIL;
-
-	switch (desc->iface) {
-	case passive_serial:
-		log_debug("Launching Passive Serial Dump\n");
-		ret_val = ACEX1K_ps_dump (desc, buf, bsize);
-		break;
-
-		/* Add new interface types here */
-
-	default:
-		printf ("%s: Unsupported interface type, %d\n",
-				__FUNCTION__, desc->iface);
-	}
-
-	return ret_val;
-}
-
-int ACEX1K_info( Altera_desc *desc )
-{
-	return FPGA_SUCCESS;
-}
-
-/* ------------------------------------------------------------------------- */
-/* ACEX1K Passive Serial Generic Implementation                                  */
-
-static int ACEX1K_ps_load(Altera_desc *desc, const void *buf, size_t bsize)
-{
-	int ret_val = FPGA_FAIL;	/* assume the worst */
-	Altera_ACEX1K_Passive_Serial_fns *fn = desc->iface_fns;
-	int i;
-
-	log_debug("start with interface functions @ 0x%p\n", fn);
-
-	if (fn) {
-		size_t bytecount = 0;
-		unsigned char *data = (unsigned char *) buf;
-		int cookie = desc->cookie;	/* make a local copy */
-		unsigned long ts;		/* timestamp */
-
-		log_debug("Function Table:\n"
-			  "ptr:\t0x%p\n"
-			  "struct: 0x%p\n"
-			  "config:\t0x%p\n"
-			  "status:\t0x%p\n"
-			  "clk:\t0x%p\n"
-			  "data:\t0x%p\n"
-			  "done:\t0x%p\n\n",
-			  &fn, fn, fn->config, fn->status,
-			  fn->clk, fn->data, fn->done);
-#ifdef CONFIG_SYS_FPGA_PROG_FEEDBACK
-		printf ("Loading FPGA Device %d...", cookie);
-#endif
-
-		/*
-		 * Run the pre configuration function if there is one.
-		 */
-		if (*fn->pre) {
-			(*fn->pre) (cookie);
-		}
-
-		/* Establish the initial state */
-		(*fn->config) (true, true, cookie);	/* Assert nCONFIG */
-
-		udelay(2);		/* T_cfg > 2us	*/
-
-		/* nSTATUS should be asserted now */
-		(*fn->done) (cookie);
-		if ( !(*fn->status) (cookie) ) {
-			puts ("** nSTATUS is not asserted.\n");
-			(*fn->abort) (cookie);
-			return FPGA_FAIL;
-		}
-
-		(*fn->config) (false, true, cookie);	/* Deassert nCONFIG */
-		udelay(2);		/* T_cf2st1 < 4us	*/
-
-		/* Wait for nSTATUS to be released (i.e. deasserted) */
-		ts = get_timer (0);		/* get current time */
-		do {
-			CFG_FPGA_DELAY ();
-			if (get_timer (ts) > CFG_SYS_FPGA_WAIT) {	/* check the time */
-				puts ("** Timeout waiting for STATUS to go high.\n");
-				(*fn->abort) (cookie);
-				return FPGA_FAIL;
-			}
-			(*fn->done) (cookie);
-		} while ((*fn->status) (cookie));
-
-		/* Get ready for the burn */
-		CFG_FPGA_DELAY ();
-
-		/* Load the data */
-		while (bytecount < bsize) {
-			unsigned char val=0;
-#ifdef CONFIG_SYS_FPGA_CHECK_CTRLC
-			if (ctrlc ()) {
-				(*fn->abort) (cookie);
-				return FPGA_FAIL;
-			}
-#endif
-			/* Altera detects an error if INIT goes low (active)
-			   while DONE is low (inactive) */
-#if 0 /* not yet implemented */
-			if ((*fn->done) (cookie) == 0 && (*fn->init) (cookie)) {
-				puts ("** CRC error during FPGA load.\n");
-				(*fn->abort) (cookie);
-				return (FPGA_FAIL);
-			}
-#endif
-			val = data [bytecount ++ ];
-			i = 8;
-			do {
-				/* Deassert the clock */
-				(*fn->clk) (false, true, cookie);
-				CFG_FPGA_DELAY ();
-				/* Write data */
-				(*fn->data) ((val & 0x01), true, cookie);
-				CFG_FPGA_DELAY ();
-				/* Assert the clock */
-				(*fn->clk) (true, true, cookie);
-				CFG_FPGA_DELAY ();
-				val >>= 1;
-				i --;
-			} while (i > 0);
-
-#ifdef CONFIG_SYS_FPGA_PROG_FEEDBACK
-			if (bytecount % (bsize / 40) == 0)
-				putc ('.');		/* let them know we are alive */
-#endif
-		}
-
-		CFG_FPGA_DELAY ();
-
-#ifdef CONFIG_SYS_FPGA_PROG_FEEDBACK
-		putc (' ');			/* terminate the dotted line */
-#endif
-
-	/*
-	 * Checking FPGA's CONF_DONE signal - correctly booted ?
-	 */
-
-	if ( ! (*fn->done) (cookie) ) {
-		puts ("** Booting failed! CONF_DONE is still deasserted.\n");
-		(*fn->abort) (cookie);
-		return (FPGA_FAIL);
-	}
-
-	/*
-	 * "DCLK must be clocked an additional 10 times fpr ACEX 1K..."
-	 */
-
-	for (i = 0; i < 12; i++) {
-		CFG_FPGA_DELAY ();
-		(*fn->clk) (true, true, cookie);	/* Assert the clock pin */
-		CFG_FPGA_DELAY ();
-		(*fn->clk) (false, true, cookie);	/* Deassert the clock pin */
-	}
-
-	ret_val = FPGA_SUCCESS;
-
-#ifdef CONFIG_SYS_FPGA_PROG_FEEDBACK
-		if (ret_val == FPGA_SUCCESS) {
-			puts ("Done.\n");
-		}
-		else {
-			puts ("Fail.\n");
-		}
-#endif
-	(*fn->post) (cookie);
-
-	} else {
-		printf ("%s: NULL Interface function table!\n", __FUNCTION__);
-	}
-
-	return ret_val;
-}
-
-static int ACEX1K_ps_dump(Altera_desc *desc, const void *buf, size_t bsize)
-{
-	/* Readback is only available through the Slave Parallel and         */
-	/* boundary-scan interfaces.                                         */
-	printf ("%s: Passive Serial Dumping is unavailable\n",
-			__FUNCTION__);
-	return FPGA_FAIL;
-}
drivers/fpga/Kconfig +0 -6
index 9456ca3149ae..18e7d7c99a9a 100644
@@ -34,12 +34,6 @@ config FPGA_STRATIX_V
  	help
  	  Say Y here to enable the Altera Stratix V FPGA specific driver.
 
-config FPGA_ACEX1K
-	bool "Enable Altera ACEX 1K driver"
-	depends on FPGA_ALTERA
-	help
-	  Say Y here to enable the Altera ACEX 1K FPGA specific driver.
-
  config FPGA_CYCLON2
  	bool "Enable Altera FPGA driver for Cyclone II"
  	depends on FPGA_ALTERA
drivers/fpga/Makefile +0 -1
index 610c168fc35c..6c779d5b5df6 100644
@@ -17,7 +17,6 @@ obj-$(CONFIG_FPGA_XILINX) += xilinx.o
  obj-$(CONFIG_FPGA_LATTICE) += ivm_core.o lattice.o
  ifdef CONFIG_FPGA_ALTERA
  obj-y += altera.o
-obj-$(CONFIG_FPGA_ACEX1K) += ACEX1K.o
  obj-$(CONFIG_FPGA_CYCLON2) += cyclon2.o
  obj-$(CONFIG_FPGA_INTEL_SDM_MAILBOX) += intel_sdm_mb.o
  obj-$(CONFIG_FPGA_STRATIX_II) += stratixII.o
drivers/fpga/altera.c +1 -4
index 64fda3a307c7..4a9aa74357ec 100644
@@ -28,10 +28,7 @@ static const struct altera_fpga {
  	int			(*dump)(Altera_desc *, const void *, size_t);
  	int			(*info)(Altera_desc *);
  } altera_fpga[] = {
-#if defined(CONFIG_FPGA_ACEX1K)
-	{ Altera_ACEX1K, "ACEX1K", ACEX1K_load, ACEX1K_dump, ACEX1K_info },
-	{ Altera_CYC2,   "ACEX1K", ACEX1K_load, ACEX1K_dump, ACEX1K_info },
-#elif defined(CONFIG_FPGA_CYCLON2)
+#if defined(CONFIG_FPGA_CYCLON2)
  	{ Altera_ACEX1K, "CycloneII", CYC2_load, CYC2_dump, CYC2_info },
  	{ Altera_CYC2,   "CycloneII", CYC2_load, CYC2_dump, CYC2_info },
  #endif
include/ACEX1K.h +0 -0
index 7c5253c66cca..44d1b9765657 100644

include/ACEX1K.h line 17

Quoted 10 lines
include/ACEX1K.h +0 -4
@@ -12,26 +12,10 @@
 
  #include <altera.h>
 
-extern int ACEX1K_load(Altera_desc *desc, const void *image, size_t size);
-extern int ACEX1K_dump(Altera_desc *desc, const void *buf, size_t bsize);
-extern int ACEX1K_info(Altera_desc *desc);
-
  extern int CYC2_load(Altera_desc *desc, const void *image, size_t size);
  extern int CYC2_dump(Altera_desc *desc, const void *buf, size_t bsize);
  extern int CYC2_info(Altera_desc *desc);

Maybe CYC2_load() etc. can be moved to a different header file like altera.h for example?

Quoted 43 lines
include/ACEX1K.h +0 -35
-/* Slave Serial Implementation function table */
-typedef struct {
-	Altera_pre_fn		pre;
-	Altera_config_fn	config;
-	Altera_clk_fn		clk;
-	Altera_status_fn	status;
-	Altera_done_fn		done;
-	Altera_data_fn		data;
-	Altera_abort_fn		abort;
-	Altera_post_fn		post;
-} Altera_ACEX1K_Passive_Serial_fns;
-
  /* Slave Serial Implementation function table */
  typedef struct {
  	Altera_pre_fn		pre;
@@ -43,37 +27,4 @@ typedef struct {
  	Altera_post_fn		post;
  } Altera_CYC2_Passive_Serial_fns;
 
-/* Device Image Sizes
- *********************************************************************/
-/* ACEX1K */
-/* FIXME: Which size do we mean?
- * Datasheet says 1337000/8=167125Bytes,
- * Filesize of an *.rbf file is 166965 Bytes
- */
-#if 0
-#define Altera_EP1K100_SIZE	1337000/8	/* 167125 Bytes */
-#endif
-#define Altera_EP1K100_SIZE	(166965*8)
-
-#define Altera_EP2C8_SIZE	247942
-#define Altera_EP2C20_SIZE	586562
-#define Altera_EP2C35_SIZE	883905
-#define Altera_EP3C5_SIZE	368011		/* .rbf size in bytes */
-
-#define ALTERA_EP4CE6_SIZE	368011		/* 2944088 Bits */
-#define ALTERA_EP4CE10_SIZE	368011		/* 2944088 Bits */
-#define ALTERA_EP4CE15_SIZE	510856		/* 4086848 Bits */
-#define ALTERA_EP4CE22_SIZE	718569		/* 5748552 Bits */
-#define ALTERA_EP4CE30_SIZE	1191788		/* 9534304 Bits */
-#define ALTERA_EP4CE40_SIZE	1191788		/* 9534304 Bits */
Quoted 8 lines

include/ACEX1K.h typedef struct { · line 71

Quoted 3 lines
include/ACEX1K.h +0 -3
-#define ALTERA_EP4CE55_SIZE	1861195		/* 14889560 Bits */
-#define ALTERA_EP4CE75_SIZE	2495719		/* 19965752 Bits */
-#define ALTERA_EP4CE115_SIZE	3571462		/* 28571696 Bits */

Can we move these definitions for device image sizes to altera.h instead removing them? I added some of those with commit 3b2a595fc659 ("fpga: altera: Add some more device sizes") back in 2019, and I'm using some of them in downstream code/dts.

I expect you are also saying that these are Cyclone 2 related right?

No issue from my side.

M

#5 lore raw

Re: [PATCH 2/3] fpga: Remove ancient ACEX1K support

Hello Michal,

Am Fri, Jul 25, 2025 at 02:17:53PM +0200 schrieb Michal Simek:

Quoted 20 lines · Alexander Dahl
Quoted 17 lines

Hello Michal,

Am Fri, Jul 25, 2025 at 01:25:53PM +0200 schrieb Michal Simek:

Quoted 14 lines

Coverity (CID 583149) reports issue on code which is not enabled by any real platform that's why remove it completely.

---

  configs/sandbox_defconfig |   1 -
  drivers/fpga/ACEX1K.c     | 245 --------------------------------------
  drivers/fpga/Kconfig      |   6 -
  drivers/fpga/Makefile     |   1 -
  drivers/fpga/altera.c     |   5 +-
  include/ACEX1K.h          |  49 --------
  6 files changed, 1 insertion(+), 306 deletions(-)

delete mode 100644 drivers/fpga/ACEX1K.c

[…]

Quoted 67 lines
Quoted 58 lines

include/ACEX1K.h line 15

Quoted 13 lines
include/ACEX1K.h +0 -4
index 7c5253c66cca..44d1b9765657 100644
@@ -12,26 +12,10 @@
  #include <altera.h>
-extern int ACEX1K_load(Altera_desc *desc, const void *image, size_t size);
-extern int ACEX1K_dump(Altera_desc *desc, const void *buf, size_t bsize);
-extern int ACEX1K_info(Altera_desc *desc);
-
  extern int CYC2_load(Altera_desc *desc, const void *image, size_t size);
  extern int CYC2_dump(Altera_desc *desc, const void *buf, size_t bsize);
  extern int CYC2_info(Altera_desc *desc);

Maybe CYC2_load() etc. can be moved to a different header file like altera.h for example?

Quoted 41 lines
include/ACEX1K.h +0 -35
-/* Slave Serial Implementation function table */
-typedef struct {
-	Altera_pre_fn		pre;
-	Altera_config_fn	config;
-	Altera_clk_fn		clk;
-	Altera_status_fn	status;
-	Altera_done_fn		done;
-	Altera_data_fn		data;
-	Altera_abort_fn		abort;
-	Altera_post_fn		post;
-} Altera_ACEX1K_Passive_Serial_fns;
-
  /* Slave Serial Implementation function table */
  typedef struct {
  	Altera_pre_fn		pre;
@@ -43,37 +27,4 @@ typedef struct {
  	Altera_post_fn		post;
  } Altera_CYC2_Passive_Serial_fns;
-/* Device Image Sizes
- *********************************************************************/
-/* ACEX1K */
-/* FIXME: Which size do we mean?
- * Datasheet says 1337000/8=167125Bytes,
- * Filesize of an *.rbf file is 166965 Bytes
- */
-#if 0
-#define Altera_EP1K100_SIZE	1337000/8	/* 167125 Bytes */
-#endif
-#define Altera_EP1K100_SIZE	(166965*8)
-
-#define Altera_EP2C8_SIZE	247942
-#define Altera_EP2C20_SIZE	586562
-#define Altera_EP2C35_SIZE	883905
-#define Altera_EP3C5_SIZE	368011		/* .rbf size in bytes */
-
-#define ALTERA_EP4CE6_SIZE	368011		/* 2944088 Bits */
-#define ALTERA_EP4CE10_SIZE	368011		/* 2944088 Bits */
-#define ALTERA_EP4CE15_SIZE	510856		/* 4086848 Bits */
-#define ALTERA_EP4CE22_SIZE	718569		/* 5748552 Bits */
-#define ALTERA_EP4CE30_SIZE	1191788		/* 9534304 Bits */
-#define ALTERA_EP4CE40_SIZE	1191788		/* 9534304 Bits */
Quoted 8 lines

include/ACEX1K.h typedef struct { · line 71

Quoted 3 lines
include/ACEX1K.h +0 -3
-#define ALTERA_EP4CE55_SIZE	1861195		/* 14889560 Bits */
-#define ALTERA_EP4CE75_SIZE	2495719		/* 19965752 Bits */
-#define ALTERA_EP4CE115_SIZE	3571462		/* 28571696 Bits */

Can we move these definitions for device image sizes to altera.h instead removing them? I added some of those with commit 3b2a595fc659 ("fpga: altera: Add some more device sizes") back in 2019, and I'm using some of them in downstream code/dts.

Quoted 1 line

I expect you are also saying that these are Cyclone 2 related right?

Yes. Only EP1K* is related to ACEX-1K. EP2C* is Cyclone II, EP3C* relates to Cyclone III, EP4C* to Cyclone IV.

These are the maximum bitstream sizes of different chip variants, and can directly be compared to .rbf file sizes (in octets). Not sure why the cyclone sizes are in that acex header, but maybe because there was no dedicated cyclone header? I would just move them to a cyclone header or to the generic altera header.

Quoted 1 line

No issue from my side.

:-)

Alex

author maintainer
#6 lore raw

PATCH 3/3 fpga: lattice: Remove unused support

In reply to Michal Simek

There is no single platform which is using this driver that's why remove it completely. Some issues regarding this code are also reported by Coverity (CID 583143, 583144, 583145, 583146).

---

 configs/sandbox_defconfig |    1 -
 drivers/fpga/Kconfig      |    6 -
 drivers/fpga/Makefile     |    1 -
 drivers/fpga/fpga.c       |   23 -
 drivers/fpga/ivm_core.c   | 3149 -------------------------------------
 drivers/fpga/lattice.c    |  379 -----
 include/fpga.h            |    1 -
 include/lattice.h         |  298 ----
 8 files changed, 3858 deletions(-)
 delete mode 100644 drivers/fpga/ivm_core.c
 delete mode 100644 drivers/fpga/lattice.c
 delete mode 100644 include/lattice.h
configs/sandbox_defconfig +0 -1
index 96546cc0bcfe..5bd26b433157 100644
@@ -207,7 +207,6 @@ CONFIG_FPGA_ALTERA=y
 CONFIG_FPGA_STRATIX_II=y
 CONFIG_FPGA_STRATIX_V=y
 CONFIG_FPGA_CYCLON2=y
-CONFIG_FPGA_LATTICE=y
 CONFIG_FPGA_XILINX=y
 CONFIG_FPGA_SPARTAN2=y
 CONFIG_FPGA_SPARTAN3=y
drivers/fpga/Kconfig +0 -6
index 18e7d7c99a9a..e2593057fac7 100644
@@ -55,12 +55,6 @@ config FPGA_INTEL_SDM_MAILBOX
 	  Enable FPGA driver for writing full bitstream into Intel FPGA
 	  devices through SDM (Secure Device Manager) Mailbox.
 
-config FPGA_LATTICE
-	bool "Enable Lattice FPGA driver"
-	help
-	  This is used for the lattice FPGAs. Please check the source code as
-	  there is no documentation for this at present.
-
 config FPGA_XILINX
 	bool "Enable Xilinx FPGA drivers"
 	select FPGA
drivers/fpga/Makefile +0 -1
index 6c779d5b5df6..f22d3b3d86e3 100644
@@ -14,7 +14,6 @@ obj-$(CONFIG_FPGA_VIRTEX2) += virtex2.o
 obj-$(CONFIG_FPGA_ZYNQPL) += zynqpl.o
 obj-$(CONFIG_FPGA_ZYNQMPPL) += zynqmppl.o
 obj-$(CONFIG_FPGA_XILINX) += xilinx.o
-obj-$(CONFIG_FPGA_LATTICE) += ivm_core.o lattice.o
 ifdef CONFIG_FPGA_ALTERA
 obj-y += altera.o
 obj-$(CONFIG_FPGA_CYCLON2) += cyclon2.o
drivers/fpga/fpga.c +0 -23
index 2297fefd149f..b4e3896d2e8a 100644
@@ -9,7 +9,6 @@
 #include <log.h>
 #include <xilinx.h>             /* xilinx specific definitions */
 #include <altera.h>             /* altera specific definitions */
-#include <lattice.h>
 #include <dm/device_compat.h>
 
 /* Local static data */
@@ -79,14 +78,6 @@ static int fpga_dev_info(int devnum)
 			ret_val = altera_info(desc->devdesc);
 #else
 			log_err("No support for Altera devices.\n");
-#endif
-			break;
-		case fpga_lattice:
-#if defined(CONFIG_FPGA_LATTICE)
-			log_info("Lattice Device\nDescriptor @ 0x%p\n", desc);
-			ret_val = lattice_info(desc->devdesc);
-#else
-			log_err("No support for Lattice devices.\n");
 #endif
 			break;
 		default:
@@ -265,13 +256,6 @@ int fpga_load(int devnum, const void *buf, size_t bsize, bitstream_type bstype,
 			ret_val = altera_load(desc->devdesc, buf, bsize);
 #else
 			log_err("No support for Altera devices.\n");
-#endif
-			break;
-		case fpga_lattice:
-#if defined(CONFIG_FPGA_LATTICE)
-			ret_val = lattice_load(desc->devdesc, buf, bsize);
-#else
-			log_err("No support for Lattice devices.\n");
 #endif
 			break;
 		default:
@@ -310,13 +294,6 @@ int fpga_dump(int devnum, const void *buf, size_t bsize)
 			ret_val = altera_dump(desc->devdesc, buf, bsize);
 #else
 			log_err("No support for Altera devices.\n");
-#endif
-			break;
-		case fpga_lattice:
-#if defined(CONFIG_FPGA_LATTICE)
-			ret_val = lattice_dump(desc->devdesc, buf, bsize);
-#else
-			log_err("No support for Lattice devices.\n");
 #endif
 			break;
 		default:
drivers/fpga/ivm_core.c deleted +0 -3149

2,927 lines not on this page · raw message

drivers/fpga/lattice.c deleted +0 -379

379 lines not on this page · raw message

include/fpga.h +0 -1
index a144238e66a1..20153b2082a5 100644

Lines of this file are not shown. View the raw message.

include/lattice.h deleted +0 -298

298 lines not on this page · raw message

The rest of this diff is not shown. View the raw message.

Signature · 1 line
2.43.0

Part of the diff is not shown here. View the raw message.

#7 lore raw
In reply to Michal Simek

Hello Michal,

Am Fri, Jul 25, 2025 at 01:25:51PM +0200 schrieb Michal Simek:

Quoted 8 lines

Hi,

when fpga drivers were enabled via sandbox some coverity issues came and make no sense to fix them when there is no user of these drivers that's why it is easier just remove them completely.

Another candidate for removal is FPGA_CYCLON2 which is enabled only by mx53cx9020_defconfig.

We use FPGA_CYCLON2 downstream with devices containing Altera Cyclone III and IV or Efinix Trion T20 chips. These are _not_ legacy devices, few quantities per year, but still produced and sold.

If you need board code to keep the upstream support, I would need to talk to some people here if we can upstream our board code, or at least some of it. That might be difficult though, so if there are alternatives, please advise. O:-)

Greets
Alex

Quoted 34 lines

Patrick: Is this board still maintained?

Thanks,
Michal

Michal Simek (3):
fpga: xilinx: Check valid desc structure
fpga: Remove ancient ACEX1K support
fpga: lattice: Remove unused support

 configs/sandbox_defconfig |    2 -
 drivers/fpga/ACEX1K.c     |  245 ---
 drivers/fpga/Kconfig      |   12 -
 drivers/fpga/Makefile     |    2 -
 drivers/fpga/altera.c     |    5 +-
 drivers/fpga/fpga.c       |   23 -
 drivers/fpga/ivm_core.c   | 3149 -------------------------------------
 drivers/fpga/lattice.c    |  379 -----
 drivers/fpga/xilinx.c     |    3 +
 include/ACEX1K.h          |   49 -
 include/fpga.h            |    1 -
 include/lattice.h         |  298 ----
 12 files changed, 4 insertions(+), 4164 deletions(-)
 delete mode 100644 drivers/fpga/ACEX1K.c
 delete mode 100644 drivers/fpga/ivm_core.c
 delete mode 100644 drivers/fpga/lattice.c
 delete mode 100644 include/lattice.h
Signature · 4 lines
2.43.0

base-commit: 021783860f7e628f7c4e0c101707cd4250e6d61f
branch: debian-sent3
author maintainer
#8 lore raw
Quoted 12 lines · Alexander Dahl

Hello Michal,

Am Fri, Jul 25, 2025 at 01:25:51PM +0200 schrieb Michal Simek:

Quoted 8 lines

Hi,

when fpga drivers were enabled via sandbox some coverity issues came and make no sense to fix them when there is no user of these drivers that's why it is easier just remove them completely.

Another candidate for removal is FPGA_CYCLON2 which is enabled only by mx53cx9020_defconfig.

Quoted 8 lines

We use FPGA_CYCLON2 downstream with devices containing Altera Cyclone III and IV or Efinix Trion T20 chips. These are _not_ legacy devices, few quantities per year, but still produced and sold.

If you need board code to keep the upstream support, I would need to talk to some people here if we can upstream our board code, or at least some of it. That might be difficult though, so if there are alternatives, please advise. O:-)

CYCLON2 is enabled in one board that's why it is fine until that board is around.

Thanks,
Michal

#9 lore raw
Quoted 19 lines · Michal Simek
Quoted 16 lines · Alexander Dahl

Hello Michal,

Am Fri, Jul 25, 2025 at 01:25:51PM +0200 schrieb Michal Simek:

Quoted 8 lines

Hi,

when fpga drivers were enabled via sandbox some coverity issues came and make no sense to fix them when there is no user of these drivers that's why it is easier just remove them completely.

Another candidate for removal is FPGA_CYCLON2 which is enabled only by mx53cx9020_defconfig.

We use FPGA_CYCLON2 downstream with devices containing Altera Cyclone III and IV or Efinix Trion T20 chips. These are _not_ legacy devices, few quantities per year, but still produced and sold.

Quoted 6 lines
Quoted 4 lines

If you need board code to keep the upstream support, I would need to talk to some people here if we can upstream our board code, or at least some of it. That might be difficult though, so if there are alternatives, please advise. O:-)

CYCLON2 is enabled in one board that's why it is fine until that board is around.

Heck, we can just leave it on for sandbox *but* that does mean someone needs to step up to maintain the driver part. The report is at
https://lore.kernel.org/u-boot/20250725132645.GA1807455@bill-the-cat/
and it's that there's some dead code. It's also been not quite 3 years since the last "real" change made to the driver by you, so are there any other general fixes / improvements that should be upstreamed? Thanks!

Signature · 1 line
Tom
#10 lore raw

Hello Tom,

Am Fri, Jul 25, 2025 at 08:00:35AM -0600 schrieb Tom Rini:

Quoted 25 lines · Michal Simek
Quoted 24 lines · Alexander Dahl
Quoted 20 lines

Hello Michal,

Am Fri, Jul 25, 2025 at 01:25:51PM +0200 schrieb Michal Simek:

Quoted 8 lines

Hi,

when fpga drivers were enabled via sandbox some coverity issues came and make no sense to fix them when there is no user of these drivers that's why it is easier just remove them completely.

Another candidate for removal is FPGA_CYCLON2 which is enabled only by mx53cx9020_defconfig.

We use FPGA_CYCLON2 downstream with devices containing Altera Cyclone III and IV or Efinix Trion T20 chips. These are _not_ legacy devices, few quantities per year, but still produced and sold.

If you need board code to keep the upstream support, I would need to talk to some people here if we can upstream our board code, or at least some of it. That might be difficult though, so if there are alternatives, please advise. O:-)

Quoted 6 lines
Quoted 1 line

CYCLON2 is enabled in one board that's why it is fine until that board is around.

Heck, we can just leave it on for sandbox *but* that does mean someone needs to step up to maintain the driver part. The report is at
https://lore.kernel.org/u-boot/20250725132645.GA1807455@bill-the-cat/
and it's that there's some dead code.

I can look into that report and at the generic fpga code, or at least some of it, maybe next week.

Quoted 3 lines

It's also been not quite 3 years
since the last "real" change made to the driver by you, so are there any other general fixes / improvements that should be upstreamed? Thanks!

Just looked into my patch stack again, there are only two things which are not board code:

1. return value wrapping in cmd/fpga.c which I submitted as a patch some years ago, was denied back then.

2. a generic dm driver for passive serial configuration over spi. upstreaming that driver would make sense together with some board code using it. ;-)

Which means: I sent patches for everything which is not board code, and most of them were accepted. :-)

Greets
Alex

#11 lore raw
In reply to Michal Simek
Quoted 5 lines · Michal Simek

Hi,

when fpga drivers were enabled via sandbox some coverity issues came and make no sense to fix them when there is no user of these drivers that's why it is easier just remove them completely.

Ah, thanks for taking care of these.

Signature · 1 line
Tom
#12 lore raw
In reply to Michal Simek

Am 2025-07-25 um 14:26 schrieb Michal Simek <[email protected]>:

Hi,
Hi Michael and Alex,

when fpga drivers were enabled via sandbox some coverity issues came and make no sense to fix them when there is no user of these drivers that's why it is easier just remove them completely.

Another candidate for removal is FPGA_CYCLON2 which is enabled only by
mx53cx9020_defconfig.
Patrick: Is this board still maintained?
The hardware yes, we still produce and sell these devices[1]. On the software side I have to be honest we didn’t touch this since beginning of 2023[2]. We upstreamed uboot to give our customers the opportunity to run their own images, so it is possible but unlikely some of them run a more recent uboot. We didn’t got much feedback from customers in the last two years, so we didn’t spend the time to add this old hardware to our modern ci. But if it helps other FPGA_CYCLON2 users in the shadows, I will try to bring this board up to date when Iam back from vacation in two weeks.

Regards,
Patrick

[1] https://www.beckhoff.com/en-en/products/ipc/embedded-pcs/cx9020-arm-r-cortex-r-a8/cx9020.html
[2] https://github.com/Beckhoff/CX9020/commit/524c74cf1efc08d4375f58157c4c831549563b70

Thanks,
Michal

…

This email contains confidential information. If you have received it in error, you must not read, use, copy or pass on this e-mail or its attachments. If you have received the e-mail in error, please inform me immediately by reply e-mail and then delete this e-mail from your system. Thank you

Diese E-Mail enthält vertrauliche Informationen. Sollten Sie sie irrtümlich erhalten haben, dürfen Sie diese E-Mail oder ihre Anhänge nicht lesen, verwenden, kopieren oder weitergeben. Sollten Sie die Mail versehentlich erhalten haben, teilen Sie mir dies bitte umgehend per Antwort-E-Mail mit und löschen Sie diese E-Mail dann aus Ihrem System. Vielen Dank

Beckhoff Automation GmbH & Co. KG | Managing Director: Dipl. Phys. Hans Beckhoff Registered office: Verl, Germany | Register court: Guetersloh HRA 7075